Wednesday, February 22, 2012

Introduction to Iowa

Remember when Dr. Paul was ahead in Iowa polling, right before the Iowa caucuses? This was a portion of the results of a survey conducted by the Des Moines Register just before the caucuses. Very interesting ... Iowa analysis coming by Friday.

Tuesday, February 21, 2012

New Hampshire update

Much has happened since my last post from 3 days ago. In the previous post, I demonstrated that in the New Hampshire Republican primary, as the number of voters per precinct increases, Mr. Romney's share of the vote increases while Dr. Paul's share of the vote decreases. I first want to address some of the questions and comments related to this post.

Comment. I can't see images or images have poor resolution
Answer. See this.

Comment. You are seeing this trend because Mr. Romney does better in urban and wealthy areas.
Answer. I cannot (yet) answer this directly, but if I can obtain county-level (or better yet, precinct-level) income data, I could address the second part of that comment. Regarding the urban Mitt-island effect, the Manchester data appears to be inconsistent with the statement, "Mr. Romney does better in urban areas." Take a look at this map of the various wards for Manchester. It appears to show that wards with smaller areas favor Dr. Paul (he won the two wards surrounded by red lines), while wards with larger areas favor Mr. Romney (he doubled Dr. Paul in the 3 wards surrounded by orange lines). Maybe someone from Manchester could specifically identify a demographic effect which would explain these rather remarkable differences in support, given the close proximity. One possible explanation is that Mr. Romney did extremely well in suburban areas. Again, it would be very nice if I could obtain precinct-level income data for New Hampshire!

Comment. Your regression fit doesn't look too good.
Answer. Whether a straight line is the best fit of the data is short-sighted. Obviously, there is a strong correlation between the difference in Mr. Romney's support and Dr. Paul's support and precinct size in 11 of the 12 counties! Statewide, the correlation is 0.9!

Moving forward with New Hampshire, I plan to analyze the data from the 2008 primary as well as look at additional precinct-level or county-level information that I have yet to collect, including population, registered Republicans, geographic area, income, property values, etc. If anyone knows where I can find some of this information, I would be thrilled to get some help.

My next few posts will take a look at Iowa, which is a different animal altogether. However, some preliminary analyses are pretty interesting. For example, when looking at counties with an average population of more than 2000 per precinct, which applies to more than half of Iowa's population, Mr. Romney received 28% of the vote while Mr. Santorum and Dr. Paul were virtually tied with 22% of the vote. In counties with less than 2000 people per precinct, Mr. Romney received only 19.5% of the vote, behind Dr. Paul with 20.3% and Mr. Santorum with 27.3%. I intend to find an explanation for this.

Saturday, February 18, 2012

New Hampshire - WOW!

I thought I would start with a state where I didn't expect to see very much. I was wrong. Let's journey through New Hampshire together, shall we?

New Hampshire is proud to be "first-in-the-nation" when it comes to primaries. Candidates who win New Hampshire have an excellent chance to win the nomination. Based on the 2010 census, there are over 1.3 million people in New Hampshire. According to the Secretary of State's website, over 300,000 people voted in this year's primary. Contrast that with neighboring Maine, whose GOP turnout so far (according to their own website) is under 6,000.

Among the candidates still running for the nomination, recall that Mr. Romney won New Hampshire decisively with 39%. Dr. Paul was 2nd with 23%, and Mr. Santorum edged out Mr. Gingrich with just under 10%.

My focus is to determine if there is a relationship between precinct size and percent of vote for that precinct. Let's first look at the distribution of all 303 precincts in terms of voter turnout.
Notice that a single precinct, because of the nature of primaries (as opposed to caucuses), can handle hundreds to thousands of voters. However, we can see that almost half of the precincts had fewer than 500 voters.

Next, I decided to take a look at some of the raw data. This graph shows the precinct-level data for Dr. Paul and Mr. Romney (I looked at the other candidates but they had uninteresting results and made the picture too messy), where each candidate's support (vertical axis) in that precinct is measured as the percentage of that precinct's total vote. I plotted that against precinct size (horizontal axis).



The straight lines represent the best linear fit of the data for each candidate (using method of least squares). I will be the first to admit that the data do not appear to be all that linear, but it's a start. If we had expected Dr. Paul to do just as well in larger precincts as he did in smaller precincts, then we certainly WOULD NOT be seeing this type of graph.

To better understand this, I decided to plot the difference between the Romney percentage of the vote and the Paul percentage of the vote (vertical axis) against precinct size (horizontal axis).

Although there is quite a lot of "noise" in the smaller precincts, there is a definite trend in Mr. Romney's favor as precinct size increases. Based on a linear relationship between these 2 variables, the correlation is 0.44 (p-value < 0.0001).

So what's the big deal, right? Maybe there are demographic differences between large and small precincts. This is certainly a possibility, and it's a question that we can directly address. I attempted to identify precincts which could be grouped based on geographic location. The largest set of precincts I found were the 12 precincts representing the 12 wards of Manchester in Hillsborough County. When you look at just these 12 precincts, one would hypothesize that the difference between Romney votes and Paul votes would not vary much based on precinct size because these 12 precincts are demographically homogeneous. I was unable to find a map showing how the different wards for Manchester are divided up, but I would assume that Ward 1 is close to Ward 2, Ward 5 is close to Ward 6, etc. Here are the 12 ward numbers, sorted by precinct size from smallest to largest: 5, 3, 11, 4, 7, 9, 10, 12, 2, 6, 8 and 1. In other words, precinct size DOES NOT appear to depend on geographic location in Manchester. Maybe someone in Manchester could better identify if demographic differences exist in the various wards of Manchester. Anyway, here's what the graph for Manchester looks like.



This is very alarming. Here is a county where Mr. Romney beat Dr. Paul by more than 18%. Manchester, which is in that same county, is the largest city in the state of New Hampshire. Inexplicably, Dr. Paul won 2 of the 3 smallest precincts in Manchester, but Mr. Romney doubled Dr. Paul's vote totals in the 3 largest precincts in Manchester.

Am I missing something?

Friday, February 17, 2012

Precinct-Level Data for IA, ME, MN, NV and NH

So far in this primary season, we have seen caucuses in Iowa, Nevada, Minnesota, Colorado and Maine. There have been primaries in New Hampshire, South Carolina, Florida and Missouri. Because these events are organized and run by each state's Republican Party leadership, there is an incredible amount of heterogeneity between the different states.

This makes things difficult for someone who wants to obtain precinct-level data for every single precinct in each of those 9 states, but that is my task. After some painstaking googling, and a decent quantity of repetitive keystrokes and mouse-clicks, I am now confident that I have ALL of the publicly available precinct-level data for Iowa, Maine, Minnesota, Nevada and New Hampshire. Finally! And yes, I know that Maine's results should be adjusted after the Feb 18 caucuses.

Okay, so I can't confirm or deny (yet) the claims of the poster at Daily Paul regarding South Carolina, but I can look at some other states and see if there's a similar trend. In other words, as precinct size increases, does Dr. Paul fare increasingly worse or Mr. Romney fare increasingly better? That is the only question I am asking for now.

Because each state has a unique set of considerations, I think it's best if I consider the results from each state separately, and I won't be able to do that until later this weekend. However, I must give a teaser: the early results appear to be VERY interesting!

For example, about 43% of the 6520 precincts from these 5 states had 20 or fewer votes. Dr. Paul won 20.5% of those precincts, while Mr. Romney only won 19.3% and Mr. Santorum won 27.2% (Mr. Gingrich won 6.7%). These precincts represent a total of 25,324 voters, or 5.5%.

Now let's consider the 10.5% of precincts with more than 100 voters. Dr. Paul won 20.1% of these precincts, Mr. Romney took a whopping 56.9%, Mr. Santorum dropped to 20.7% and Mr. Gingrich captured a paltry 1.5%. These precincts represent an incredible 318,741 voters, or 69.9%.

Hmm ... just a guess, but I think that the Santorum camp may find this preliminary data to be fairly interesting.

My first state-specific post will deal with New Hampshire, and the graph below is certainly intriguing. It shows the relationship between vote percentage and precinct size for both Dr. Paul and Mr. Romney in New Hampshire.





South Carolina GOP primary - questions raised

A few months ago, the Wall Street Journal published an article about the recent party elections in Russia. Because Putin's ruling Russia United party appeared to fare particularly well in precincts with high voter turnout, 100,000 Russians marched against Putin over the controversy.
















Could something similar be happening in the GOP primaries and caucuses right here in America? Surely not ... right?!

Yesterday (Feb 17), I ran across a posting at Daily Paul, which as a disclaimer is run by Ron Paul supporters. The poster suggested that something similar to what was seen in Putin's Russia may have happened in the South Carolina GOP primary last month. This person had been studying the Greenville County precinct-level election data (available here) and found something rather unusual: "Ron Paul averaged 24% in precincts where less than 250 people voted; he averaged less than 12 percent in precincts with more than 800." Dr. Paul would have captured about 20% of the Greenville County vote based on the smallest 35 precincts but ended up with only 17.2%. The opposite trend was seen when looking at Mr. Romney's vote; that is, he fared much better on a percentage basis in the larger precincts than in the smaller ones. The poster suggested that it is possible that someone could have taken votes from Dr. Paul and given those votes to Mr. Romney, possibly accomplished via manipulation of the software installed on the voting machines.

This study, which I found very interesting, raised a number of questions in my mind, and I think I know what the questions are. I also think I know in what order they must be answered:
  1. Is there convincing empirical evidence, backed up by sound statistical reasoning, that Mr. Romney's percentage of the vote in a precinct is directly proportional to that precinct's size (One could ask a similar question regarding Dr. Paul)? This question must obviously be asked of statewide results and not just in South Carolina. This is the topic of my next post.
  2. If the answer to #1 is an irrefutable "Yes," then is there a mechanism (other than election fraud) which can sufficiently explain these effects? It has been suggested that, because older voters tend to favor Mr. Romney over Dr. Paul, these voters may also live in disproportionately larger precincts. Another possible explanation is that smaller precincts are more rural and thus Dr. Paul may poll better in rural areas. I would submit that it will be very difficult to show that the answer to this question is clearly "No."
  3. However, if the answer to #2 is "No," and we have truly exhausted all other possible explanations for this phenomenon, it is at this point that we must consider jumping down the rabbit hole to explore worlds unknown. I sincerely hope that this line of questioning will not lead me to be asking questions such as, "Who is that man behind the curtain?"
I am concerned that definitively answering the first question above may take several weeks, but I will pursue it until I am certain of the answer.

Do we have fair elections in America?

This is a question that I find myself asking with increasing frequency these days. So I decided to organize my thoughts here.

When people ask the question above, they immediately think of rigging elections directly; that is, by stuffing ballot boxes with forged ballots, "counting" the votes in a way that favors a certain candidate, or by controlling electronic voting machines. But there are other ways to affect the outcomes of elections that could be construed as "unfair":
  • Voter intimidation or coercion: The democrats proposed a bill a few years ago to allow unions to be created via card check, and opponents correctly argued that this would eliminate a worker's right to a secret ballot and thus be pressured by others to vote in favor of union creation. And who can forget the presence of some members of the New Black Panther Party stationed outside a polling location on the day Mr. Obama was elected President?
  • Polls can influence opinions (rather than report them): Whether it's opinion polls, favorability ratings, or exit/entrance polls, there is plenty of opportunity for improvement. Some common problems with polls include: poorly designed studies in which samples are anything but representative, biased interpretation of results, and influencing how stories are reported in the media.
  • The media: Certainly it's possible for the media to give positive press to candidates or ideas that their bosses agree with, and vice versa (can you say Ron Paul).
  • The government: Some would argue that the government has some influence on the media and thus indirectly have some influence on elections. I concur with that reasoning.
  • Money: Does money help to win elections? Or do the attractive candidates raise the most money? Definitely worth investigating, given the fundraising surges seen with the Santorum campaign after Iowa and the Gingrich campaign after South Carolina.
I will be investigating these kinds of topics as well as some fundamental topics on the current state of the political and economic systems, both here in the United States and around the world. Maybe some of you will join me in this quest for truth.